CRO/Sponsor

Clinical Trial Supply Chain

Decentralized Clinical Trials

Early Phase Development

Investigative Sites

Protocol Design

Risk-Based Monitoring

Study Start-Up

Trial Design

Clinical Operations & Strategy

Diversity Action Plans

Industry Initiatives

Recruitment

Analytics

Electronic Data Capture

Metrics & Benchmarks

Real World Evidence

Data Management

CRAACO

Enrollment

Integrated Research

Patient Participation

Institutional Review Board & Ethics Committee

Regulatory

Artificial Intelligence/Machine Learning

eClinical

mHealth

Technology

News

Academy

Publications

All Publications

E-Books

Media

Editorial Podcasts

Editorial Videos

Peer Exchange

Practice Academy

Conferences

Conference Coverage

Conference Listing

Columns

All Columns

A Closing Thought

Clinical Trial Insights

View from Brussels

View from Washington

Resources

Advertise

Blogs

Events

Front & Center

Partner Perspectives

White Papers

Webcasts

Subscribe

Latest solutions, case studies, interviews, practical articles for clinical trials operations, data management, technology, investigative sites and services.

About

Contact Us

Editorial Contacts

Do Not Sell My Personal Information

Privacy Policy

Terms and Conditions

Applied Clinical Trials Online

LinkedIN

Instagram

Facebook

Youtube

Soundcloud

Threads

Spotlight

DE&I

Topic

1rem

Authors

This article will examine the ways to detect fraud and data fabrication within a clinical trial.

How to Detect, Manage, and Report Fraud and Fabricated Clinical Research Data

Video interview with Sepehr Shojaei, VP, Design Solutions, Lightship

Lightship's Approach to Accessible, Representative, and Impactful Clinical Trials

Lightship prioritizes patient accessibility and centricity in clinical trials, ensuring a seamless and positive experience for every participant. Learn more about:

Lightship's mission and core purpose as a company

How Lightship’s mobile nursing teams and home visits enhance study participation and retention

The role of mobile research units in the community and how they enhance participant education, accessibility, and convenience

Sepehr Shojaei, VP of Design Solutions at Lightship, talks about how the company works with pharma and contract research organization (CRO) partners to support clinical trials that are patient-first and accessible.

Latest Conference Coverage

Lightship’s Approach to Accessible, Representative, and Impactful Clinical Trials

Video interview with David MacMurchy, CEO of Lightship

Accelerate Inclusive Enrollment and Delivery with Lightship

Lightship focuses on patient accessibility and centricity in clinical trials, ensuring the patient experience is a positive one. Learn more about:

Lightship's impact on achieving diversity in clinical trials by engaging underrepresented patient populations

How Lightship utilizes hybrid and decentralized trial designs to reduce patient burden

The impact a patient centric model has on clinical trial participation and outcomes

David MacMurchy, Chief Executive Officer at Lightship, offers some insights into the company's mission toward improving equity and diversity in clinical trials. 

In this video interview, Rob Jones product manager, TMF practice area, Pharmalex, discusses challenges the TMF space is currently facing and how it can utilize collaboration in making advancements. 

Battling Outdated Regulations in the Trial Master File Space

, Rob Jones, product manager, TMF practice area, Pharmalex, discussed challenges in the trial master file (TMF) space, highlighting outdated regulations such as ICH E6(R3). When it comes to innovation and compliance in the space, he stressed the importance of a risk-based approach, focusing on critical records and leveraging artificial intelligence for big data analysis. Moving forward, Jones believes the TMF community's collaborative nature can be viewed as a strength in navigating regulatory changes and technological advancements.

ACT: What are some challenges the TMF space is currently facing? How do you think they can be addressed?

It's a nice, broad question. I think really the biggest challenge is keeping up. I think right now we've got new regulations coming in, things like ICH E6(R3) have finally come in this year, after being delayed for a little bit because of other things that were going on in the early 2020s. We've got new proposed standards from the reference model with V4 and all of this while trying to balance that desire for innovation, cost effective solutions, just an ever-changing landscape. I think one of the things that we have to look at and admit is regulations have been a little out of step for a while. We were expecting ICH E6(R3) months, if not years ago, but again, due to things that got delayed. That's fine, but a lot of things have already passed it by. The EMA has already put in new guidelines around their perception of the TMF. They had a brilliant new guideline around computerized system validation, which included things like the electronic TMF. The MHRA have been giving us more and more feedback on their current public forums and Q&As and things like that. It's kind of a little too late in some ways, ICH E6(R2) came out in 2016, and even then it was only a small look there in the grand scheme of things, realistically, it was written back in 1996. I don't need to tell you how much has changed since 1996. My marker for a lot of these things when I talk with people, is the first iPhone came out in 2007, so anything post-2007, you have had constant satellite communication world in your pocket. We're an industry that's built on stability, order, organization, we're managing clinical trial documents, but this new world hasn't been that for a long time, it's messy in a lot of ways.

The challenge really is, what do we do when the regulations are a little bit too old and outdated? How do we manage when other industries, especially outside of clinical trials, are just flying by, and what can we do to keep up without jeopardizing people's comfort, without jeopardizing all of the hard work that's gone into it? I think the nice thing really, with the TMF, and this shows every single time I go to a conference. It is one of the very few parts of clinical trials where it's an open book. There is a real community behind everything. Everyone's talking to each other. You go and talk with people, and they will tell you that TMF stands for “too much fun.” They'll tell you that TMF stands for “too many friends.” There is really this idea of everyone's in it together. I think the nice thing with that is that's been built from the beginning. The reference model when it came up was sponsors, CROs, vendors, consultants, all working together, sharing what they were doing. Nothing was secret. Nothing was hidden behind closed doors. I think, really, that is the biggest strength, and that's how we can address this ever-changing landscape, if we all come to get together and make a bit of a collective decision and say, “We as a group are going to make an interpretation of this, and this is how we're going to take it forwards.” There's a little bit where perception becomes reality, and if we're all pushing in that same direction, that really helps with how we can address all of this. I think, again, just for the TMF world, that's really how we're going to get by with these things changing at such a rate. It's one of the biggest strengths of our little, niche part of the industry.

In this video interview, Rob Jones, product manager, TMF practice area, Pharmalex, discusses challenges the TMF space is currently facing and how it can utilize collaboration in making advancements. 

Awareness of mental health disorders has significantly improved in recent decades in the US. Due to this elevated awareness there has been an increase in diagnoses, opening the door to new clinical interventions. According to the National Institute of Health, 22.8% of Americans over the age of 18 experience a diagnosable mental health disorder in any given year. Anxiety disorder affects 19.1%, while 8.3% suffer from depressive illnesses annually. Schizophrenia and psychotic disorders are rarer (0.25%-0.64%), however they can have an outsized impact on society. Schizophrenia is one of the top 15 leading causes of disability globally. No matter the diagnosis, mental health disorders affect the quality of life of the patient and their families.

Opportunities of SaMD use to treat mental health

In parallel with greater awareness and diagnoses, there have been major technological advances resulting in increased processing power in smaller devices. Digital health is transforming the healthcare industry, with software aiding the diagnosis and treatment of illnesses. Software as a medical device (SaMD) enhances patient and healthcare provider access to real-time health information, enabling earlier diagnosis and access to a broader range of treatment options. Treating mental health disorders using SaMD is being studied in controlled research environments for indications including major depressive disorders and schizophrenia. The development of smartphone applications is an encouraging new avenue to treat these conditions.

Challenges of integrating SaMD in clinical trials

As with developing any new treatment, there are challenges when integrating these technologies into clinical trials. Any potential risks and benefits must be weighed carefully, as they are in other clinical trials. SaMD as a treatment for mental health disorders has additional challenges related to the target patient population. Issues of consent, ethical considerations and patient recruitment need to be handled with special care due to the vulnerabilities of the patients.

For SaMDs to treat mild mental health disorders the issue of mental capacity and trial informed consent can proceed as in a traditional clinical trial. However, some of the patients who potentially have the most to gain from using SaMD have more severe mental health disorders impacting their mental capacity to provide informed consent. A legally authorized representative may be necessary to provide informed consent for trial participation. Sponsors must implement standardized assessments of patient mental capacity to give consent. There are several acceptable scales which can be used to screen potential trial participants, including the Clinical Global Impression and World Health Organization Disability Assessment Schedule 2.0. Disorder-specific scales can also be employed to assess the risk and benefit of a clinical trial inclusion for each patient. Furthermore, it is important to incorporate scales that assess patient mental stability and safety, such as the Columbia Suicide Severity Rating Scale.

Patients with compromised mental health often have cognitive impairments and may be overwhelmed by SaMD technology. These patients may struggle with adoption and/or adherence to SaMD treatment requirements. Thus, a lack of human factor/usability understanding can negatively impact the value of SaMD technologies to clinical trials affecting patient outcomes and the quality, validity, and reliability of data. During SaMD design and development, the patients’ technological capabilities to comply with the SaMD treatment requirements (abilities to understand and use a smartphone, tablet or laptop application), need to be assessed. This assessment should be strongly considered (at a minimum) during the development of the SaMD as the understanding of human factors and usability risks will directly impact the efficacy of the treatment. For trials incorporating already available SaMDs, prior to enrolment and randomization, standardized assessments can provide a uniform evaluation across multiple sites to mitigate the risk of the impact of technology adoption and adherence on trial success.

Data is essential to the success of any clinical trial. When performing clinical trials of SaMD digital therapeutics, the device used must be capable of tracking user data per patient to ensure protocol compliance. The data gathered by the trial application version is likely to be more comprehensive than the original product design specifications to meet trial primary end points. Equally importantly, the developer must take extra measures to ensure that patient protected health information is safeguarded. This can be done with a comprehensive review of SaMD access restrictions, access reviews, and retention term across the sponsor’s integrated systems for data storage and processing.

Sponsors need a multifaceted approach to training and technical support for the SaMD trial. No one team can be responsible for both the training and technical support required for a clinical trial of the SaMD digital therapeutic. A collaborative, proactive approach by the sponsor and clinical research organization (CRO) will ensure that all sites and patients are appropriately trained. A common avoidable mistake is expecting the clinical team to troubleshoot technical issues with the SaMD when they occur. Appropriate expert technical support should be available to the CRO, sites, investigators, and/or patients. The sponsor and CRO should proactively anticipate potential risks and put mitigation measures in place.

In all clinical trials the ability to optimize and predict enrolment by the target patient population is a critical factor. SaMD studies introduce additional variables which are not present in other studies. Collaboration between the sponsor and a CRO to identify multiple potential sites and principal investigators (PI) is an important first step to addressing this issue. Targeting a single site “type” for a given SaMD protocol places an unnecessary risk if the target site/PI criteria are wrong for this patient pool. Sponsor and CRO collaboration to evaluate a diverse set of sites for inclusion can mitigate this risk. Seamless, clear communication between site staff, sponsor, and the CRO will ensure that feedback and recommendations are actioned. To prevent patient dropout and the risk of timeline extensions, stipends for participation and reimbursement of transport costs should be considered.

The level of regulatory control is dependent on the risk level associated with the healthcare decisions made in response to the data provided by the software application. The FDA regulates an app as SaMD when it makes claims regarding the diagnosis or treatment of a mental health condition. Clinical data will likely be required to support the safety and effectiveness of these SaMDs. The clinical evaluation will closely examine SaMD inputs, algorithms, and outputs via clinical association validation, analytical/technical validation, and clinical validation. Sponsors must provide evidence demonstrating the SaMD’s ability to deliver clinically meaningful and accurate results relevant to its intended use. Clinical validation should confirm that the SaMD fulfils its intended purpose within the target population and the specific clinical care context for which it was designed.

SaMDs have the potential to improve the lives of patients managing mental health disorders, by providing accessible, scalable, patient-centered treatment options. Clinical trials will always have unexpected challenges and SaMD trials are no exception. These studies require additional measures to address ethical and consent considerations due to a vulnerable patient population. Additional technical and regulatory support through partnerships with a CRO allows sponsors greater adaptability to changing circumstances. Close collaboration between the sponsor and CRO can optimize patient recruitment, study budget, study timelines, and pathway to market for a new or enhanced SaMD.

, PhD, director, project management, medical device and diagnostics research; and 

, BSCS, RAC, senior manager, regulatory affairs; both with ICON

With close collaboration between sponsors, CROs, and regulators, this emerging technology has the potential to greatly improve the lives of patients.

Clinical Strategies to Optimize Software as a Medical Device to Treat Mental Health

View the Applied Clinical Trials April 2025 issue in an interactive format.

Applied Clinical Trials: April 2025 Interactive Digital Edition

Oncology clinical trials are increasingly complex, driven by rising data volumes, regulatory requirements, and inefficiencies in manual data handling. Current trial processes require research staff to manually extract, transcribe, and validate clinical data from electronic health records (EHRs) into electronic data capture (EDC) systems. These manual processes, such as data entry, consume significant time and resources. More than 50% of clinical trial data is duplicated between research systems and hospital EHRs, with around 20% of total study costs typically allocated to data duplication and verification. As oncology trials expand to incorporate genomic, biomarker, imaging, and digital health data, the volume of data per patient is multiplying, intensifying site burden, increasing costs, and slowing trial execution.

eSource technology offers a transformative solution by enabling automated EHR-to-EDC data transfers, eliminating redundant data entry, and improving accuracy. By seamlessly integrating structured data formats such as FHIR, HL7, SNOMED, and LOINC, eSource ensures real-time data availability, reduces errors, and enhances regulatory compliance. Additionally, artificial intelligence (AI)-powered automation is increasingly being leveraged to process unstructured data, such as physician notes, imaging, and pathology reports, further reducing site workload and improving data standardization.

Early adopters of eSource, such as Memorial Sloan Kettering, Mayo Clinic, and MD Anderson Cancer Center, have reported significant improvements in trial efficiency. Case studies show that eSource can reduce site burden by 50%, cut transcription errors, and accelerate database lock times, ultimately shortening the time to bring new therapies to market.

As oncology research centers scale eSource adoption, the field is shifting toward a fully digital, clinical trial ecosystem. The transition to real-time, interoperable data capture not only increases efficiency, but also supports adaptive trial designs, enhances patient recruitment, and fosters a collaborative ecosystem among sponsors, technology vendors, and research institutions. eSource represents a pivotal advancement in the future of oncology trials, ensuring they are faster, more cost-effective, and data-driven.

Overview of participating oncology research centers

The six oncology centers that are part of this study represent a broad spectrum of expertise, size, and specialties, playing pivotal roles in global cancer research and care. These centers are also recognized as eSource Champion sites, actively running or preparing for eSource-enabled trials and pioneering innovations in clinical trial efficiency. Mayo Clinic Comprehensive Cancer Center (MCC), spanning three US sites, excels in translational research and conducts more than 1,500 trials annually, supported by 500+ research staff.

City of Hope National Medical Center (COH) in California emphasizes innovative oncology treatments, enrolling 5,000 patients in 730+ annual trials and digitizing workflows for efficiency. MD Anderson Cancer Center (MDA) in Houston, Texas, one of the world’s largest, has 760 inpatient beds and conducts 1,600+ trials yearly. Memorial Sloan Kettering Cancer Center (MSK) in New York, a global leader in cancer care and research, features 514 inpatient beds and currently has more than 1,300 therapeutic clinical trials open, excelling in advanced therapies and precision oncology.

The Cancer Research UK Cambridge Center, including Cambridge University Hospitals (CUH)NHS Foundation Trust, specializes in investigator-initiated trials and translational science within a university hospital, leveraging structured data integration. University Hospital of Essen (UKE) in Germany, with around 1,300 beds, is a leader in digitization, using FHIR-based data systems and AI-driven solutions for unstructured data. Together, these institutions span from 500 to 1,300+ beds and conduct hundreds to thousands of trials annually, underscoring their significant contributions to advancing cancer research and patient care globally. As eSource Champion hospitals, these centers exemplify the potential of digital transformation in oncology research, setting benchmarks for operational excellence and collaboration.

This study employed a qualitative research design to explore the adoption, implementation, and scaling of eSource technology across oncology research centers. Semi-structured interviews were conducted with senior representatives, who also served as co-authors, from the six leading institutions: MDA, MCC, COH, CUH, UKE, and MCK. These respondents offered comprehensive insights into the utilization and impact of eSource systems based on their institutional expertise.

The interview guide was structured to capture experiences with current eSource implementations, perceived benefits, technical and operational challenges, strategies for managing unstructured data, and collaborative practices among stakeholders. Each 60-minute interview was audio-recorded, transcribed, and subjected to thematic analysis to identify recurring themes, including interoperability, emerging trends in eSource, broader benefits to stakeholders, and operational efficiency.

To ensure validity, preliminary findings were shared with participants for feedback, facilitating refinements and confirmation of interpretations. The study further assessed the validity and generalizability of the interview outcomes, finding them consistent with the operational practices and strategic priorities of the participating institutions. This iterative and collaborative process reinforced the credibility of the results and ensured alignment with institutional realities.

By synthesizing diverse perspectives, this study illuminates critical success factors for scaling eSource technology in oncology trials and offers actionable insights to improve data integration, streamline workflows, and enhance trial efficiency.

Current use of eSource technology in oncology

The adoption of eSource technology across leading oncology research centers has significantly enhanced data management, trial efficiency, and workflow optimization. By integrating EHR-to-EDC automation, institutions are reducing site burden, improving data accuracy, and enabling real-time access to critical trial data.

 Several oncology research centers have successfully implemented or are in the process of adopting eSource-enabled trials, demonstrating measurable improvements. At MCC, the deployment of an eSource system has centralized data entry and streamlined workflows, allowing research staff to focus on high-value tasks. Angela Fritsche emphasized, “EHR-to-EDC allows staff to focus on higher-level tasks and improves transparency for patients and sponsors.” Similarly, COH has integrated eSource technology, significantly reducing transcription errors and manual workload. Matthew Bush highlighted, “We reduced transcription errors entirely and decreased data transcription times from 15 minutes to under two minutes per subject.”

 eSource technology has proven its ability to scale across multicenter trials. MDA pilot studies demonstrate its feasibility in high-volume research environments, with Lee Andrews stating, “Our coordinators were happy with the time saved from a manual abstraction perspective.” At CUH, eSource has automated lab data transfer, reducing errors and ensuring seamless integration with trial databases. Sarah Burges emphasized, “We want to get rid of the dull, error-prone activity of cutting and pasting. …It’s nonsense in this day and age.”

Beyond efficiency gains, eSource improves data integrity and ensures compliance with General Data Protection Regulation (GDPR) and HIPAA regulations. Automated data transfer reduces discrepancies, minimizing protocol deviations, and enhancing audit readiness. With early adopters demonstrating tangible benefits, broader adoption is expected to reshape oncology clinical trials, making them more efficient, scalable, and data driven.

EHR interoperability and data quality interoperability remains a key challenge in scaling eSource technology for oncology trials. While FHIR, HL7, SNOMED CT, and LOINC provide structured data frameworks, adoption varies across institutions, requiring extensive mapping and validation. Felix Nensa from UKE noted, “We have close to 500 source systems—mapping these to FHIR repositories requires continuous updates and fine-tuning.” Fritsche from MCC emphasized the need for early data dictionary mapping to ensure reliable extraction. Overcoming these inconsistencies demands collaboration between hospitals, regulatory bodies, and vendors to standardize oncology trial data and enable seamless automation.

Despite its benefits, eSource implementation faces hurdles in interoperability, data mapping, and legal compliance. Fritsche described the resource-intensive process of mapping the data dictionary but noted, “The transition was smooth with clear expectations.” Andrews from MDA cited legal and data privacy reviews as obstacles, while Nensa from UKE highlighted delays in adapting smart-on-FHIR for non-Epic systems.

Frequent protocol amendments and unstructured data further complicate operations. Bush from COH questioned, “How do we maintain these studies without excessive resource demands?” Addressing these challenges is essential for fully realizing eSource’s potential.

Change management and adoption readiness. 

Implementing eSource technology requires more than technical upgrades—it demands organizational change management to overcome resistance and streamline adoption. Many research teams are used to manual data entry, making the transition to automation challenging. Burges from CUH noted, “The trust in automation was immediate—after a week, the research staff felt comfortable, but initial skepticism had to be addressed.” Training gaps and compliance protocols also pose barriers, while bureaucracy and legal reviews delay adoption. Andrews from MDA highlighted, “Legal and data privacy concerns have significantly slowed down vendor onboarding.” Targeted training and stakeholder engagement are essential for successful eSource integration.

Beyond three minutes: Unveiling the true time and cost of oncology data management. 

A retrospective analysis from the EIT Health-sponsored EHR2EDC project (2018–2020) by AstraZeneca and Sanofi found that study personnel spend at least three minutes per data point in Phase II-III oncology trials.

 This includes searching, manual entry, review, and query resolution. For a study with 10 patients contributing 10,000 data points each, the cumulative workload reaches 5,000 hours. In Phase III trials with 200+ patients, the time and costs increase exponentially, highlighting the urgent need for automation and eSource adoption.

Insights from MSK 2024 eSource use case suggest that the three-minute estimate for data entry is conservative for standard data points.

 Complex entries, such as adverse events or abnormal lab values, can take five to 10 minutes per entry. Input from five additional oncology centers corroborates these findings, with respondents reporting that straightforward data points require five minutes, while complex entries demand even more time.

Bush from COH noted, “The variability of data complexity makes a significant difference—what used to take three minutes can now take upwards of 10, depending on the query’s depth.”

Nensa from UKE emphasized the challenge, stating, “Every additional layer of complexity in the data expands the time needed to ensure accuracy and compliance.

These revised estimates carry major cost implications. In a trial with 10,000 data points per patient, increasing the average entry time from three to five minutes raises the total workload by 33%—equating to $30,000 to $50,000 in transcription costs per patient and 8,300 hours of additional effort for 10 patients. However, eSource adoption results in $15,000 in cost savings per patient, alleviating resource burdens (see Figure 1).

Scaling across Phase III trials, manual entry requires thousands of additional hours, significantly inflating costs. eSource technology offers a transformative solution by automating data capture and integrating AI-driven workflows. MCC has reported a 50% reduction in data entry times with eSource implementation. Fritsche emphasized, “Automation is critical—not just for speed but for accuracy and scalability.” These advancements enhance efficiency, improve data integrity, and accelerate clinical research, ensuring that oncology trials are faster, more cost-effective, and scalable.

Unstructured data, including clinical notes, imaging, and pathology reports, remains a major challenge in eSource implementation. Nensa from UKE highlighted AI’s potential, stating, “AI-driven solutions are already capable of transforming text into standardized formats.” However, variability in physician notetaking and imaging structures complicates automation. To address this, strategies include creating structured fields within EHRs to reduce reliance on free text. Burgess noted, “The challenge with unstructured data is that every site documents things differently. Standardizing EHR fields for key oncology data points would eliminate much of the manual effort needed for data abstraction.”

The adoption of eSource technology presents transformative opportunities for oncology clinical trials, enhancing efficiency, data quality, and collaboration. By automating data capture, reducing manual site burden, and enabling AI-driven structured data integration, eSource accelerates research while improving accuracy and compliance. Additionally, eSource facilitates the improved design of electronic case report forms, ensuring that data collection aligns seamlessly with study protocols, enhances standardization, and reduces the need for post-hoc data reconciliation.

Reducing site burden through automation. 

A significant advantage of eSource is its ability to automate data capture and eliminate redundant manual transcription. Traditionally, clinical research coordinators spend substantial time manually extracting and transcribing patient data from EHRs to EDC systems. By integrating FHIR, HL7, and standardized application programming interfaces, eSource automates this process, reducing transcription errors and allowing research staff to focus on higher-value tasks. Fritsche from MCC emphasized, “EHR-to-EDC automation enables staff to focus on patient care instead of being bogged down by data entry.” This shift improves workflow efficiency and reduces site workload.

Enhancing data quality through AI integration. AI-driven eSource solutions significantly improve data accuracy, completeness, and regulatory compliance. AI-powered natural language processing (NLP) and machine learning enable structured data extraction, ensuring consistency across multi-site trials. MSK has reported a 99% reduction in data entry errors after implementing eSource.

 Bush from COH highlighted, “AI-driven automation has the potential to improve data integrity, reducing discrepancies that arise from human input.” Improved data quality not only accelerates trial timelines but also enhances regulatory submissions and real-world evidence generation.

Collaborative ecosystem development for standardization.

 Scaling eSource requires multi-site, multi-vendor collaborations to drive interoperability and industry-wide standardization. Stakeholder alignment among hospitals, sponsors, regulatory agencies, and technology vendors is critical for ensuring widespread adoption.

Burges from CUH stressed, “Collaboration is the tipping point—sponsors must mandate eSource to drive industry-wide adoption.” Establishing pre-competitive frameworks and cross-industry alliances will facilitate seamless data sharing and reduce implementation barriers. By reducing burden, improving data quality, and fostering collaboration, eSource unlocks new possibilities for oncology research, enabling faster, more cost-effective, and patient-centered clinical trials.

As the complexity of oncology trials grows, AI is emerging as a key enabler in eSource implementation, streamlining data extraction, improving trial efficiency, and enhancing decision-making. By automating manual processes, structuring unstructured data, and supporting predictive analytics, AI transforms how clinical trials manage and utilize real-world data.

 AI-powered NLP and machine learning reduce the manual workload required for data abstraction. Traditionally, research staff spend hours transcribing EHR data into EDC systems, but AI-driven automation extracts, maps, and validates structured and unstructured data in real time. Nensa from UKE noted, “We no longer need human intervention for routine data abstraction—AI algorithms can accurately extract and map patient data directly to the trial database.” This accelerates trial timelines and allows research teams to focus on more complex tasks.

AI in trial design, patient stratification, and safety monitoring.

 Beyond automating data transfer, AI optimizes trial design, patient recruitment, and risk management. AI-driven algorithms analyze large datasets to identify ideal patient cohorts, improving trial enrollment. AI also enhances adverse event detection, monitoring clinical data in real time to flag safety risks before they escalate. Fritsche from MCC stated, “AI-driven risk detection models allow us to intervene earlier in patient care, reducing the likelihood of severe adverse events.”

Unstructured data—such as physician notes, imaging scans, and pathology reports—poses a significant challenge in oncology trials. AI-powered tools convert these complex datasets into structured, analyzable formats. At MSK, AI processes radiology and pathology reports, extracting key biomarkers. Andrews from MDA highlighted, “By leveraging AI, we can integrate imaging and genomic data directly into trial workflows, reducing manual effort in data extraction.”

Navigating regulatory uncertainty in AI in clinical trials.

 Despite its transformative potential, the adoption of AI in clinical research faces challenges, particularly around regulatory acceptance and compliance. While regulatory agencies have signaled support for AI’s ability to improve trial efficiency, they have not defined clear, standardized pathways for validating its use. This uncertainty leaves clinical trial sponsors cautious, balancing the promise of faster, more accurate data collection with the risk of regulatory pushback during submissions. As the regulatory framework evolves, collaborative efforts between research sites, sponsors, technology providers, and the regulators will be crucial in implementing AI-driven clinical trial workflows.

The next two to five years will be transformative for eSource technology in oncology clinical trials, driving efficiency, scalability, and real-time insights. Experts anticipate a fundamental shift toward automated data integration, real-time analytics, and AI-powered decision-making, establishing eSource as the standard method for clinical trial data management.

The future of eSource in clinical trials.

 Respondents widely foresee a future where eSource is fully embedded into trial workflows, reducing manual data entry and enabling more efficient study execution. Burges from CUH remarked, “We’ll see eSource becoming the backbone of all trials, reducing manual input and enabling faster, more accurate insights.”

This transition is expected to bring near real-time data flow, minimizing delays between data collection and analysis. AI-driven automation will also play a critical role in eliminating inefficiencies in unstructured data processing. Fritsche from MCC noted, “In three years, we’ll laugh at discussing structured data. The future is unstructured data and AI learning disease-specific languages.” This signals a major shift from structured, rule-based data extraction toward AI-driven, adaptive learning models that can interpret physician notes, imaging, and genomics.

AI and machine learning as key enablers. 

Emerging technologies such as AI and machine learning are set to revolutionize patient recruitment, trial matching, and risk monitoring. Bush from COH emphasized AI’s role in clinical decision-making, stating, “AI could transform adverse-event signaling and structured tumor measurements.” AI’s ability to process large datasets in real time will enhance adverse event detection, predictive modeling, and trial optimization. Nensa from UKE underscored the potential of AI-driven text-to-FHIR models such as the “Pigeon” system, which converts unstructured text into standardized FHIR-formatted data. This innovation could significantly improve data interoperability and accessibility across oncology research centers.

The transformational shift: Fully eSource-enabled trials.

 The transition to fully eSource-enabled studies marks a pivotal moment in clinical research, where all participating sites leverage automated data capture, real-time integration, and AI-driven validation. This shift eliminates manual data transcription, enhances interoperability, and enables real-time trial monitoring, significantly reducing the administrative burden on research sites. Within the next two years, Phase I oncology trials are expected to be fully eSource-enabled, leading to a six-month reduction in database lock times. The impact on cost efficiency, trial speed, and data accuracy will be profound, helping to bring new therapies to market faster.

As illustrated in Figure 2 below, this transformation replaces traditional clinical trial execution—characterized by manual data entry, fragmented systems, and inefficiencies—with a streamlined, AI-powered ecosystem. eSource reduces site burden, improves data integrity, and enhances regulatory compliance, creating a scalable, interoperable framework for clinical research.

By integrating real-time data capture, AI-driven validation, and standardized data structures, eSource ensures trials are conducted more efficiently, cost-effectively, and with greater precision. This evolution not only optimizes trial workflows and compliance but also positions oncology research centers at the forefront of digital transformation in clinical trials.

Key strategies for eSource implementation

Scaling eSource technology requires a multi-faceted approach that balances operational readiness, stakeholder engagement, and regulatory alignment. A common theme among respondents was the need for collaboration between sponsors, healthcare organizations, and technology vendors to drive adoption and standardization.

Early adopters highlight strategic planning, resource allocation, and interoperability as essential for successful eSource implementation. Andrews from MDA emphasized, “It’s incumbent upon major players to push the industry to standardize around eSource technology.” MCC’s approach included proactive data mapping and system integration, with Fritsche noting, “The upfront work of mapping the data dictionary was a lift, but the transition was smooth with clear expectations.”

To scale eSource effectively, organizations must prioritize structured data formats, compliance frameworks, and cross-site collaboration. Burges from CUH stressed, “Sponsors need to mandate eSource use to create a tipping point for broader adoption.”

Successful eSource adoption requires alignment among research sites, sponsors, and technology providers. Bush from COH emphasized, “Collaboration across sites is key to creating scalable, replicable processes.” Early integration planning ensures EHR-to-EDC interoperability from the start, while regulatory alignment with HIPAA, GDPR, and FDA 21 CFR Part 11 is critical for compliance. Additionally, training and change management help research staff adopt eSource technology effectively, reducing resistance and maximizing efficiency in clinical trial workflows.

Ensuring interoperability and data flow. 

Interoperability remains a key challenge in scaling eSource, requiring standardized data structures and seamless system integration. Oncology research centers face EHR variability and vendor restrictions, complicating automation. Nensa from UKE noted, “Once the data is standardized and structured, we’ve completed 80% of the journey.” Adopting FHIR, HL7, SNOMED, and LOINC ensures real-time data exchange, improves consistency, and enhances trial efficiency and compliance.

Early adopter case studies/collaboration. 

Institutions leading eSource adoption emphasize the value of real-world case studies to demonstrate feasibility and impact. eSource has streamlined workflows, reduced manual entry, and improved trial efficiency, allowing research staff to focus on patient care. Andrews from MDA expressed this industry shift succinctly: “Thank God, I can finally focus on talking to patients rather than being a data monkey.”

Pre-competitive collaboration is also essential. Burges advocated for “multi-vendor, multi-sponsor collaborations” to tackle interoperability challenges. Andrews reinforced this sentiment: “The pendulum swings back to balance when collaboration drives alignment across the ecosystem.”

The adoption of eSource technology is transforming clinical trials, enhancing efficiency, data quality, and scalability. Oncology has led this shift, but other therapeutic areas, including neurology, cardiology, and rare diseases, are also embracing eSource to streamline research and reduce manual burdens. A key driver of this expansion is the i-HD eSource Scale-Up Task Force,

 which plays an impartial role in accelerating adoption across sites, sponsors, and vendors. By fostering collaboration and standardization, it helps ensure seamless integration and scalability across the clinical trial ecosystem.

As eSource becomes the standard for clinical data management, it will enable real-time data flow, adaptive trial designs, and improved patient engagement. Continued collaboration among stakeholders will be critical.

The European Institute for Innovation through Health Data, i-HD; 

Cambridge University Hospitals NHS Foundation Trust; 

Memorial Sloan Kettering Cancer Center (MSK)

1. Kalankesh, L.R.; Monaghesh, E. Utilization of EHRs for Clinical Trials: A Systematic Review.

https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-024-02177-7

2. Sundgren, M.; Santiago, G.; Lengfellner, J. Streamlining Clinical Trials with eSource: Insights from MSK. 

https://www.appliedclinicaltrialsonline.com/view/streamlining-clinical-trials-with-esource-insights-from-msk

3. Hume, S.; Abolafia, J.; Low, G. Use of HL7 FHIR as eSource to Pre-populate CDASH Case Report Forms Using a CDISC ODM API. CDISC. 

4. Lengfellner, J.; Yeatman, R. EHR-to-EDC at Memorial Sloan Kettering Cancer Center (MSK). Podium presentation. March 3, 2025. Association for Clinical Data Management, Prague, Czech Republic.

5. Participating research centers in study: 

7. Ammour, N.; Griffon, N.; Djadi-Prat, J.; et al. TransFAIR Study: A European Multicenter Experimental Comparison of EHR2EDC Technology to the Usual Manual Method for eCRF Data Collection. 

The challenges, opportunities, and strategic outlook for oncology research centers. 

Lysa Triantafillou

Articles

Latest Updated Articles